Bug 79889 - Patch to update net/nvnet port to latest version
Summary: Patch to update net/nvnet port to latest version
Status: Closed FIXED
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: Normal Affects Only Me
Assignee: Pav Lucistnik
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-04-14 01:50 UTC by Quinton Dolan
Modified: 2005-04-21 08:39 UTC (History)
0 users

See Also:


Attachments
file.diff (4.53 KB, patch)
2005-04-14 01:50 UTC, Quinton Dolan
no flags Details | Diff
nvnet.diff (6.27 KB, patch)
2005-04-14 01:50 UTC, Quinton Dolan
no flags Details | Diff
smime.p7s (2.31 KB, application/pkcs7-signature)
2005-04-20 00:43 UTC, Quinton Dolan
no flags Details
smime.p7s (2.31 KB, application/pkcs7-signature)
2005-04-20 14:56 UTC, Quinton Dolan
no flags Details
nvnet.diff.2 (4.54 KB, application/octet-stream)
2005-04-20 16:42 UTC, Gregg Cooper
no flags Details
smime.p7s (2.31 KB, application/pkcs7-signature)
2005-04-21 00:35 UTC, Quinton Dolan
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Quinton Dolan 2005-04-14 01:50:20 UTC
This is an update to the net/nvnet port to bring it up to the latest version

the following files have been removed:
nvnet/distinfo.amd64
nvnet/distinfo.i386
nvnet/files/patch-src_if__nv.c
nvnet/files

the following files have been added:
nvnet/distinfo

Fix: the following is a patch to update the port:
Comment 1 Pav Lucistnik freebsd_committer freebsd_triage 2005-04-18 19:12:15 UTC
State Changed
From-To: open->feedback

On my i386/5.4-STABLE box it fails with 

===>  Building for nvnet-1.0.0301 
"Makefile", line 7: Malformed conditional (${OSTYPE} == "FreeBSD") 
"Makefile", line 7: Need an operator 
"Makefile", line 13: if-less endif 
"Makefile", line 13: Need an operator 
make: fatal errors encountered -- cannot continue 
*** Error code 1 

Looks like != assignments are not processed when running from ports' Makefile. 
cd work/nvnet && make works. : 


Comment 2 Pav Lucistnik freebsd_committer freebsd_triage 2005-04-18 19:12:15 UTC
Responsible Changed
From-To: freebsd-ports-bugs->pav

Watch
Comment 3 Gregg Cooper 2005-04-19 19:23:47 UTC
pav,

a non-root user compiling the pre-patched port yields the same:

===>  Building for nvnet-1.0.0274
"Makefile", line 7: Malformed conditional (${OSTYPE} == "FreeBSD")
"Makefile", line 7: Need an operator
"Makefile", line 13: if-less endif
"Makefile", line 13: Need an operator
make: fatal errors encountered -- cannot continue
*** Error code 1

this is not new behavior introduced by the patch ... is there a way to fix
the port? haven't found much documentation on how to do ports that install
kernel modules, and it appears there are many different methods currently in
use.

in either case, compiling the port (patched or not) works when run as root in
multi-user mode (single user mode also yields the above) ...

gregg
Comment 4 Pav Lucistnik freebsd_committer freebsd_triage 2005-04-19 19:35:10 UTC
Gregg Cooper pí¹e v út 19. 04. 2005 v 14:23 -0400:
> pav,
> 
> a non-root user compiling the pre-patched port yields the same:
> 
> ===>  Building for nvnet-1.0.0274
> "Makefile", line 7: Malformed conditional (${OSTYPE} == "FreeBSD")
> "Makefile", line 7: Need an operator
> "Makefile", line 13: if-less endif
> "Makefile", line 13: Need an operator
> make: fatal errors encountered -- cannot continue
> *** Error code 1
> 
> this is not new behavior introduced by the patch ... is there a way to fix
> the port? haven't found much documentation on how to do ports that install
> kernel modules, and it appears there are many different methods currently in
> use.


We are on something (me and maintainer), see the Audit-Trail in the PR.
I believe it's unrelated to UID of user launching compilation.

-- 
Pav Lucistnik <pav@oook.cz>
              <pav@FreeBSD.org>

A two-eyed cyclops would be a bicyclops.
Comment 5 Quinton Dolan 2005-04-20 00:43:14 UTC
On 20/04/2005, at 4:23 AM, Gregg Cooper wrote:

> pav,
>
> a non-root user compiling the pre-patched port yields the same:
>
> ===>  Building for nvnet-1.0.0274
> "Makefile", line 7: Malformed conditional (${OSTYPE} == "FreeBSD")
> "Makefile", line 7: Need an operator
> "Makefile", line 13: if-less endif
> "Makefile", line 13: Need an operator
> make: fatal errors encountered -- cannot continue
> *** Error code 1
>
> this is not new behavior introduced by the patch ... is there a way to 
> fix
> the port? haven't found much documentation on how to do ports that 
> install
> kernel modules, and it appears there are many different methods 
> currently in
> use.
>
> in either case, compiling the port (patched or not) works when run as 
> root in
> multi-user mode (single user mode also yields the above) ...

I can fix the port from generating this error when built as a non root 
user, but root privs will still be required to actually install it 
obviously.

-- 
Seeya...Q

                -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

                           _____  /  Quinton Dolan - q@OntheNet.com.au.
   __  __/  /   /   __/   /      /      OntheNet - Internet Provider
      /    __  /   _/    /      /        Gold Coast, QLD, Australia
   __/  __/ __/ ____/   /   -  /            Ph: +61 419 729 806
                     _______  /
                             _\

Comment 6 Gregg Cooper 2005-04-20 14:15:24 UTC
--- Pav Lucistnik <pav@FreeBSD.org> wrote:
> I believe it's unrelated to UID of user launching compilation.

agreed. it appears to be related to the launching shell (csh vrs sh & bash).
the port make is happy if the non-root user is using csh instead of sh or
bash.

any idea as to what in the world of make that csh is able to handle/ignore,
but gets stuck in the gears of sh and bash?

gregg
Comment 7 Quinton Dolan 2005-04-20 14:56:49 UTC
On 20/04/2005, at 11:15 PM, Gregg Cooper wrote:

> --- Pav Lucistnik <pav@FreeBSD.org> wrote:
>> I believe it's unrelated to UID of user launching compilation.
>
> agreed. it appears to be related to the launching shell (csh vrs sh & 
> bash).
> the port make is happy if the non-root user is using csh instead of sh 
> or
> bash.
>
> any idea as to what in the world of make that csh is able to 
> handle/ignore,
> but gets stuck in the gears of sh and bash?

There isn't really anything wrong with sh/bash, tcsh just defines 
OSTYPE as part of it's environment. It's only exposed because of a bug 
in my makefile.inc, I am only setting OSTYPE if OSVERSION is not 
defined. When compiled as a port OSVERSION is set by build environment 
but not OSTYPE, hence the error.

-- 
Seeya...Q

                -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

                           _____  /  Quinton Dolan - q@OntheNet.com.au.
   __  __/  /   /   __/   /      /      OntheNet - Internet Provider
      /    __  /   _/    /      /        Gold Coast, QLD, Australia
   __/  __/ __/ ____/   /   -  /            Ph: +61 419 729 806
                     _______  /
                             _\

Comment 8 Gregg Cooper 2005-04-20 16:42:41 UTC
--- Q <q@onthenet.com.au> wrote:
> There isn't really anything wrong with sh/bash, tcsh just defines 
> OSTYPE as part of it's environment. It's only exposed because of a bug 
> in my makefile.inc, I am only setting OSTYPE if OSVERSION is not 
> defined. When compiled as a port OSVERSION is set by build environment 
> but not OSTYPE, hence the error.

_thank_you_ for the explanation! without things being spelled out, i wasn't
seeing the source of the problem.

adding the following line to the port Makefile fixes the error, and is used
by other ports in the tree:

MAKE_ENV=     OSTYPE=${OPSYS}

the updated diff is attached.

gregg
Comment 9 Pav Lucistnik freebsd_committer freebsd_triage 2005-04-20 16:46:06 UTC
Gregg Cooper pí¹e v st 20. 04. 2005 v 11:42 -0400:
> --- Q <q@onthenet.com.au> wrote:
> > There isn't really anything wrong with sh/bash, tcsh just defines 
> > OSTYPE as part of it's environment. It's only exposed because of a bug 
> > in my makefile.inc, I am only setting OSTYPE if OSVERSION is not 
> > defined. When compiled as a port OSVERSION is set by build environment 
> > but not OSTYPE, hence the error.
> 
> _thank_you_ for the explanation! without things being spelled out, i wasn't
> seeing the source of the problem.
> 
> adding the following line to the port Makefile fixes the error, and is used
> by other ports in the tree:
> 
> MAKE_ENV=     OSTYPE=${OPSYS}


Yes this works. You're ok with it, Q?

-- 
Pav Lucistnik <pav@oook.cz>
              <pav@FreeBSD.org>

Can't sing. Can't dance. Can handle a sword a little.
Comment 10 Quinton Dolan 2005-04-21 00:35:22 UTC
On 21/04/2005, at 1:46 AM, Pav Lucistnik wrote:

> Gregg Cooper píe v st 20. 04. 2005 v 11:42 -0400:
>> --- Q <q@onthenet.com.au> wrote:
>>> There isn't really anything wrong with sh/bash, tcsh just defines
>>> OSTYPE as part of it's environment. It's only exposed because of a 
>>> bug
>>> in my makefile.inc, I am only setting OSTYPE if OSVERSION is not
>>> defined. When compiled as a port OSVERSION is set by build 
>>> environment
>>> but not OSTYPE, hence the error.
>>
>> _thank_you_ for the explanation! without things being spelled out, i 
>> wasn't
>> seeing the source of the problem.
>>
>> adding the following line to the port Makefile fixes the error, and 
>> is used
>> by other ports in the tree:
>>
>> MAKE_ENV=     OSTYPE=${OPSYS}
>
> Yes this works. You're ok with it, Q?



Yes, this is fine. I will actually release a new version of the driver 
shortly to fix this properly and a couple of other minor things, but 
this will do nicely in the meantime ;)

-- 
Seeya...Q

                -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

                           _____  /  Quinton Dolan - q@OntheNet.com.au.
   __  __/  /   /   __/   /      /      OntheNet - Internet Provider
      /    __  /   _/    /      /        Gold Coast, QLD, Australia
   __/  __/ __/ ____/   /   -  /            Ph: +61 419 729 806
                     _______  /
                             _\

Comment 11 Pav Lucistnik freebsd_committer freebsd_triage 2005-04-21 08:39:07 UTC
State Changed
From-To: feedback->closed

Committed, thanks!